基于卷积神经网络的裸体图片识别摘要卷积神经网络是近年来广泛应用于模式识别、图像处理等领域的一种高效识别算法，它具有结构简单、训练参数少和适应性强等特点。本文从卷积神经网络的发展历史开始，详细阐述了卷积神经网络的网络结构、神经元模型和训练算法在此基础上以卷积神经网络在裸体图片识别和形状识别方面的应用为例，简单介绍了卷积神经网络在工程上的应用，并给出了设计思路和网络结构。关键字：模型；结构；训练算法；裸体图片识别；形状识别AbstractConvolutionneuralnetworkisanefficientrecognitionalgorithmwhichiswidelyusedinpatternrecognition,imageprocessingandotherfieldsrecentyears.Ithasasimplestructure,fewtrainingparametersandgoodadaptabilityandotheradvantages.Inthispaper,beginwiththehistoryofconvolutionalneuralnetworks,describesthestructureofconvolutionalneuralnetwork,neuronmodelsandtrainingalgorithmsindetail.Onthisbasis,usestheapplicationsofconvolutionalneuralnetworkinfacedetectionandshaperecognitionasexamples,introducestheapplicationsofconvolutionneuralnetworkinengineering,andgivesdesignideasandnetworkstructure.Keywords：Model;TrainingAlgorithm;Advantage;Facedetection;Shaperecognition目录摘要..................................................................................................................................................1Abstract.............................................................................................................................................21引言..............................................................................................................................................41.1卷积神经网络的发展历史.................................................................................................42卷积神经网络...............................................................................................................................52.1网络结构.............................................................................................................................52.2神经元模型.........................................................................................................................72.3卷积网络的训练过程.........................................................................................................92.4卷积神经网络的优点.......................................................................................................123卷积神经网络的应用..................................................................................................................123.1基于卷积网络的形状识别...............................................................................................123.2基于卷积网络的裸体图片识别.......................................................................................143.2.1遮挡灵敏度............................................................................................................163.2.2去卷积网络模型....................................................................................................174总结............................................................................................................................................21参考文献........................................................................................................................................22附录...........................................................................................
